Projection of future spatial distributions of marine populations is a central question for ecologists and managers. The measure of projection uncertainty is particularly important because projections can only be useful if they are given with a known, and sufficiently high level of confidence. Uncertainties can arise for the observation process, conceptual and numerical model formulations, parameter estimates, model evaluation, appropriate consideration of spatial and temporal scales and finally the potential for adaptation of living systems. Comprehensive analyses of these multiple sources of uncertainty have not been carried out so far and how these uncertainties are considered in current studies has not yet been described. To analyse how these different sources of uncertainty are currently considered in marine research, we performed a survey of published literature during the period 2005-2009. From the 75 publications selected, we calculated how frequently each type of uncertainty was considered. We found that little attention is given to most sources of uncertainty, except for uncertainty in parameter estimates. As a result, most current projections are expected to be far less reliable than usually assumed. We conclude that, unless uncertainty can be better accounted for, such projections may be of limited use, or even risky to use for management purpose.